DE102019216054B4 - Method for monitoring injectors of an internal combustion engine and internal combustion engine for carrying out such a method - Google Patents

Abstract

Method for monitoring injectors (5) of an internal combustion engine (1) with a plurality of combustion chambers (3), each combustion chamber (3) being assigned at least one injector (5), with the following steps: a) Acquisition of combustion chamber-specific exhaust gas temperature values for at least one Group of combustion chambers (3) of the internal combustion engine (1); b) Formation of an exhaust gas temperature mean value (Tm) of the exhaust gas temperature values recorded in step a) over the group of combustion chambers (3); c) Calculation of normalized exhaust gas temperature values (Tnorm) by the in step a ) recorded exhaust gas temperature values are offset individually with the exhaust gas temperature mean value (Tm) formed in step b); d) multiple repetitions of steps a) to c) within a predetermined time interval; e) formation of pairs of values by, for each repetition of steps a) to c ) the normalized exhaust gas temperature values (Tnorm) formed in step c) in each case the average exhaust gas temperature value recorded in step b) of the same repetition ( Tm) is assigned; f) formation of value pair groups of the value pairs for each combustion chamber (3) of the internal combustion engine (1), and g) evaluation of the value pair groups according to at least one metric and obtaining at least one measure per value pair group.

Description

The invention relates to a method for monitoring injectors of an internal combustion engine and an internal combustion engine which is set up to carry out such a method.

Injector malfunctions are a frequent cause of malfunctions in the operation of internal combustion engines. Injectors which are set up for introducing fuel into the combustion chambers of internal combustion engines, in particular for direct injection or direct injection, are complicated mechanical and hydraulic systems which can have very different malfunctions. The actual cause of such a malfunction can lie in the injector itself, but external causes such as, for example, contamination of a fuel injected by the injector, detached particles from other defective components, for example flow control valves, and the like come into question. If an injector fails, this can lead to the corresponding internal combustion engine no longer achieving its rated output. An injector that doses too high or too little fuel leads to unevenly distributed combustion between different combustion chambers, which is therefore not optimal in terms of consumption or emissions. Thermal problems can also occur in the outlet area and / or in the rest of the exhaust system, including undefined temperatures of individual exhaust gas flows. As a result, effective exhaust gas aftertreatment and thus emissions can be negatively influenced in a sensitive manner. Therefore, injectors that show a malfunction must be replaced within short time intervals.

It has been proposed to automatically monitor the exhaust gas temperature of each individual combustion chamber in order to monitor injectors of an internal combustion engine. An alarm can be set if the exhaust gas temperature exceeds a predetermined limit value. There is a risk of false positive alarm messages if the exhaust gas temperature is high anyway due to the current operating point. Furthermore, the alarm occurs only comparatively late, in particular when a fault in an affected injector is already clearly pronounced. If the predetermined limit value is selected to be more sensitive, current environmental conditions and the current power of the internal combustion engine must be taken into account. This is only possible with difficulty, since assumptions have to be made that are highly error-prone, especially for an estimation of the instantaneous power. Furthermore, the alarm described here cannot offer any possibility of observing the behavior of individual injectors and describing their development. The procedure also offers no way of identifying the type of error that has occurred. The alarm is either - if the limit value is less sensitive - relatively late and requires a relatively short-term reaction, often putting the internal combustion engine out of operation, or - if the limit value is more sensitive - it is unspecific and draws a chain of precautionary service measures or further restrictions of the error.

Out EP 3 299 588 A1 a method for damage detection during the operation of an internal combustion engine emerges in which a mean value of individual measured values is determined over a specified sampling period of an ensemble of temperature sensors in or on the internal combustion engine, individual temperature differences between the mean value and the individual temperature measurement values being calculated over the specified sampling period , the individual temperature differences being calculated for successive sampling periods over a predetermined time interval. A first distribution is created by dividing the temperature differences assigned to a temperature sensor into temperature difference intervals for the predefined time interval. The first distribution is compared with a second distribution of temperature differences likewise divided into temperature difference intervals, and an operating signal is generated on the basis of a negative result of the comparison.

The invention is based on the object of creating a method for monitoring injectors of an internal combustion engine and an internal combustion engine for carrying out such a method, the disadvantages mentioned being at least reduced, preferably avoided.

The object is achieved in that the present technical teaching is provided, in particular the teaching of the independent claims and the embodiments disclosed in the dependent claims and the description.

The object is achieved in particular by creating a method for monitoring injectors of an internal combustion engine, the internal combustion engine having a plurality of combustion chambers, with at least one injector being assigned to each combustion chamber, the method comprising the following steps: recorded for a group of combustion chambers of the internal combustion engine. It b) an exhaust gas temperature mean value of the exhaust gas temperature values detected in step a) is formed over the group of combustion chambers. C) normalized exhaust gas temperature values are calculated in that the exhaust gas temperature values recorded in step a) are individually offset against the average exhaust gas temperature value formed in step b). Steps a) to c) are repeated d) several times within a predetermined time interval. E) pairs of values are formed in that, for each repetition of steps a) to c), the exhaust gas temperature values recorded in step a) are each assigned the average exhaust gas temperature value recorded in step b) of the same repetition. F) value pair groups of the value pairs are formed for each combustion chamber, and the value pair groups are g) evaluated according to at least one metric, with at least one measure per value pair group being obtained. It has been found that the application of at least one metric to the value pair groups defined here as well as the at least one measure allow statements to be made even about small previous damage to injectors, which in particular can lead to significant changes in the value pair groups. In this way, malfunctions of the injectors can be predicted and addressed at an early stage. An observation of the temporal development of the value pair groups, in particular according to the at least one metric, in particular the at least one measure, also allows an observation of the course of damage and possibly a temporal prognosis. The value pair groups also allow a statement to be made about the type of damage or an error, in particular using the at least one metric and obtaining the at least one measure. In particular, various errors / damages lead to characteristic changes in the value pair groups. The method proposed here is also advantageously insensitive to variable environmental conditions, since these are eliminated by the normalizations in step c). In particular, an influence of the outside temperature can thus be eliminated.

Damage to injectors can advantageously be detected early with the aid of the method proposed here, and its course can be observed, ideally even predicted. Various types of injector damage can be differentiated at an early stage. Since the method only uses exhaust gas temperature value measurements, it does not require any changes to the existing hardware of known internal combustion engines. The software necessary for the evaluation according to the method described here can be implemented in existing systems without further ado - possibly even without renewed certification - or the method can be used by remote maintenance on internal combustion engine data transmitted online, whereby no local implementation of the method is required. The data from numerous internal combustion engines can then be evaluated centrally, and the operators of the internal combustion engines can be informed when an injector fault is identified. At the same time, large amounts of data can be processed and meaningful statistics can be created, which in turn can improve the method proposed here and in particular the pattern recognition in the value pair groups formed. Injector faults can thus be diagnosed even earlier and more reliably, in particular if the method is used centrally on a large number of internal combustion engines over a longer period of time.

Preferably, exactly one injector is assigned to each combustion chamber of the internal combustion engine. The method then allows statements to be made about this one injector. If, on the other hand, a plurality of injectors is assigned to a combustion chamber, the method proposed here allows a statement to be made about the injector collective assigned to the combustion chamber. A statement on individual injectors of a plurality of injectors of a single combustion chamber, on the other hand, is at least not readily possible.

The injectors are preferably set up to introduce fuel into the respectively assigned combustion chamber, in particular to introduce the fuel directly into the respective combustion chamber, preferably to inject it.

The exhaust gas temperature values specific to the combustion chamber are preferably recorded at a respective outlet of the respective combustion chamber. It is also possible that the combustion chamber-specific exhaust gas temperature values are recorded on an exhaust manifold in the area of the respectively assigned combustion chamber.

The fact that the combustion chamber-specific exhaust gas temperature values are recorded in step a) for a group of combustion chambers of the internal combustion engine preferably means that they are recorded for a plurality of combustion chambers whose exhaust gas is discharged together along the same exhaust gas piping. If the internal combustion engine has a plurality of combustion chamber groups, in particular a plurality of cylinder banks, the exhaust gas temperature values specific to the combustion chamber are preferably recorded jointly for such a combustion chamber group, in particular a cylinder bank. The method is then preferably carried out separately for each of the combustion chamber groups or cylinder banks individually for each of the combustion chamber groups or cylinder banks. The internal combustion engine can be designed, for example, as a V-motor or as a W-motor. If, on the other hand, the internal combustion engine has only a single combustion chamber group comprising all combustion chambers, if it is designed, for example, as an in-line engine, then the group of combustion chambers according to step a) preferably includes all combustion chambers of the internal combustion engine, i.e. exhaust gas temperature values for each combustion chamber of the internal combustion engine are recorded .

The combustion chamber-specific exhaust gas temperature values are recorded in step a) - during one and the same iteration of steps a) to c) - in particular in the same working cycle of the internal combustion engine, i.e. within two revolutions of the crankshaft, i.e. within a crankshaft angle interval of 720 °.

The fact that in step c) the exhaust gas temperature values recorded in step a) are offset individually with the exhaust gas temperature mean value formed in step b) means in particular that the recorded exhaust gas temperature values are individually related to the exhaust gas temperature mean value, in particular in such a way that influencing variables that are relevant together for all exhaust gas temperature values such as Environmental conditions, for example an outside temperature, for which further processes are excluded or irrelevant. Such a normalization of the exhaust gas temperature values can be carried out according to a preferred embodiment in that each individual exhaust gas temperature value is divided by the average exhaust gas temperature value. In a particularly preferred embodiment, however, a difference is formed between each individually recorded exhaust gas temperature value and the average exhaust gas temperature value, with the average exhaust gas temperature value being subtracted from the individual exhaust gas temperature value. The normalized exhaust gas temperature value calculated in this way thus has a positive sign if the exhaust gas temperature exceeds the mean exhaust gas temperature value, and it has a negative sign if the detected exhaust gas temperature value is lower than the mean exhaust gas temperature value. In any case, a standardized exhaust gas temperature value is formed in step c) for each exhaust gas temperature value recorded.

In step d), steps a) to c) are repeated in particular for a plurality of work cycles within the predetermined time interval. The work cycles can follow one another directly; however, it is also possible that between two work cycles for which steps a) to c) are carried out, other work cycles are arranged for which steps a) to c) are not carried out.

The fact that value pair groups are formed in step f) means in particular that the value pairs are classified. A particularly simple embodiment of the formation of value pair groups consists in dividing the average exhaust gas temperature values into specific areas, with pairs of values with corresponding average exhaust gas temperature values then being assigned to the corresponding areas. However, more complex classification methods are preferred. The value pair groups are formed in particular on the basis of a grouping criterion or classification criterion, with those value pairs of a specific combustion chamber that belong together with regard to the grouping criterion or classification criterion being sorted into the same value pair group. Such a grouping or classification criterion can be, for example, a value pair density in a plane that is spanned by the two sizes of the value pairs, namely the normalized exhaust gas temperature on the one hand and the average exhaust gas temperature value on the other. For example, all value pairs are classified in the same value pair group, at the location of which the value pair density is greater than or equal to a certain density threshold value, a boundary of the value pair group running where the value pair density falls below the density threshold value.

In step g), in particular the at least one metric is applied to the value pair groups. Since the value pair groups of the value pairs are formed in a combustion-chamber-specific manner for each individual combustion chamber of the internal combustion engine, in step g) the at least one measure per value pair group and per combustion chamber is obtained.

An assessment of the at least one injector then preferably takes place on the basis of the at least one metric, in particular on the basis of the at least one dimension figure. As an alternative or in addition, the at least one injector is preferably assessed on the basis of the occurrence or absence of specific value pair groups. This is because it has also been found that the presence or absence of specific value pair groups in specific areas of the previously defined level is characteristic of specific damage and / or faults in an injector.

According to a further development of the invention it is provided that h) steps a) to g) are repeated for a plurality of predetermined time intervals. In this way, in particular, a change in the value pair groups, in particular in the at least one metric, in particular in the at least one measure per value pair group, can be observed. Damage or faults in injectors can thus be observed in their course. A prognosis for the further development of the injector is particularly preferably possible.

According to a further development of the invention, it is preferably provided that i) the value pair groups formed in step f) for different time intervals are assigned to one another individually in the combustion chamber. Thus, the data observed in the different time intervals for a combustion chamber are semantically linked with one another, whereby the assignment of the value pair groups enables an observation of a course or a development of individual value pair groups over time. It turned out that the value pair groups typically change their shape, center of gravity or center position, and / or extent in the previously defined plane over time. It is therefore necessary in the data collected at different times to distinguish between independent value pair groups, which may only occur randomly in the same area of the level, and data belonging together, which are to be assigned to the same value pair group. For this purpose, for example, the development over time of the value pair groups can be observed in data records that are collected in immediate succession, where value pair groups recorded in similar areas of the level are typically to be assigned to one another. If these migrate within the level over the course of time, this can be tracked accordingly, whereby a misallocation in particular of value pair groups migrated into a certain area at a later point in time to value pair groups arranged in this area in the past can be avoided.

Preferably, j) the measures of value pair groups assigned to one another, obtained in step g) in the various time intervals, are compared, and at least one comparison value per value pair group - and in particular per combustion chamber - is obtained. In this way, a change in the corresponding value pair groups can be quantified. The at least one injector is preferably evaluated on the basis of the at least one comparison value. In particular, the comparison value allows a statement to be made about the development of the at least one injector over time, and possibly a prognosis for its further development.

According to a further development of the invention, it is provided that the functionality of the at least one injector is evaluated for each combustion chamber of the internal combustion engine by comparing the at least one measure with a measure threshold value. This represents a particularly simple and at the same time reliable way of evaluating the at least one injector.

Alternatively or additionally, the functionality of the at least one injector is preferably assessed for each combustion chamber of the internal combustion engine by comparing the at least one comparison value with a comparison value threshold value. This also enables a simple and reliable assessment of the functionality of the at least one injector and, moreover, the inclusion of the time component for the purpose of analyzing a development of the functionality and / or predicting a future development of the same.

A comparison of the measure with a measure threshold value and / or the comparison value with a comparison value threshold value also enables the generation of various alarms which indicate certain injector damage - possibly for the future. A degree of previous damage to an injector can also be estimated and, if necessary, a failure probability can be derived.

According to a further development of the invention, it is provided that a development over time of the at least one measure and / or the at least one comparison value is recorded, with the respective at least one injector being evaluated on the basis of the recorded development over time for each combustion chamber. This enables in a special way the monitoring of changes in the functionality of the injectors, as well as the detection of a degree of previous damage as well as an extrapolation of a damage course and thus a prediction for the future, in particular the derivation of a failure probability and / or a predicted remaining service life. In this way, suitable measures can be identified and initiated at an early stage, and failures and / or maintenance can be better planned.

According to a further development of the invention, it is provided that a prognosis of the functionality of the at least one injector is determined for each combustion chamber on the basis of an extrapolation of the recorded development over time. This enables a statement to be made in a special way about future damage, failure probabilities and, in particular, remaining lifetimes. In this way, it is possible to identify suitable measures at an early stage and to better plan outages and / or maintenance.

According to a further development of the invention, it is provided that a value pair density in the plane spanned by the normalized exhaust gas temperature and the average exhaust gas temperature is used as the grouping criterion for the formation of the value pair groups in step f). As already explained above, in this way value pair groups can be formed in a particularly simple and reliable manner on the basis of a meaningful and easily testable criterion.

In particular, a density threshold value is used as a criterion for group membership, as already explained above.

According to a development of the invention it is provided that the at least one metric is selected from a group consisting of: a center of gravity, in particular a weighted center of gravity; a central position, in particular a geometric center position; an expansion in at least one direction, in particular in the direction of the normalized exhaust gas temperature; a position of a maximum of a value pair group, in particular a maximum normalized exhaust gas temperature; a value pair density within the value pair group; and a number of the value pairs in the value pair group. Each of these metrics is preferably based on the currently considered value pair group. The metrics listed here are particularly capable of making a statement about the value pair group under consideration. In particular, based on at least one such metric, the development over time of a corresponding value pair group can be tracked in a meaningful and meaningful way, from which in turn statements about the functionality of an injector can be obtained. In particular, even small previous damage to an injector typically leads to significant changes in the metrics listed here, for example to a significant increase in the expansion of a value pair group in the direction of the standardized exhaust gas temperature or to a shift in the center of gravity. Previous damage to an injector can be proven weeks and sometimes even months before the classic alarm routines are triggered. In particular, different metrics change in a specific way for different injector damage, for example a shift and an increase in a value pair group towards higher values along the axis of the normalized exhaust gas temperature is characteristic of damage to the seat of a pilot needle of the injector. An occurrence of value pairs with a very low value pair density in a range of high values on the axis of the normalized exhaust gas temperature with a simultaneous very large expansion of the value pair group can be characteristic of a pilot needle of an injector jammed in its guide.

According to a further development of the invention, it is provided that only exhaust gas temperature values or standardized exhaust gas temperature values are used, when they are detected, steady-state operation of the internal combustion engine is present within a predetermined steady-state time interval. The exhaust gas temperature values and / or the standardized exhaust gas temperature values are accordingly preferably suitably selected or filtered accordingly. The filtering takes place in particular in such a way that only those values are further taken into account for which at least one secondary condition is met that indicates or ensures steady-state operation of the internal combustion engine. In particular, only those exhaust gas temperature values or standardized exhaust gas temperature values are used for the evaluation in which a speed of the internal combustion engine is constant for the predetermined steady-state time interval, i.e. in particular within the predetermined steady-state time interval - especially when viewed in the past - not more than 10 revolutions per minute (rpm) has deviated upwards or downwards from a current value. Alternatively or additionally, it is preferably checked whether an individual exhaust gas temperature of a combustion chamber under consideration is constant within the predetermined steady-state time interval - viewed in the past - i.e. not more than 1 ° C up or down from a current value in the last 10 seconds, for example has deviated. The fact that the exhaust gas temperature values or standardized exhaust gas temperature values are filtered by means of a steady state condition ensures that dynamic transition effects between different operating points of the internal combustion engine, which are physically very complicated to describe and whose effects on the individual exhaust gas temperatures are therefore difficult to interpret, are excluded from consideration . Furthermore, environmental conditions - which are eliminated by the standardization - do not need to be taken into account in this consideration, and consideration of the current power of the internal combustion engine - which can become very imprecise in the event of a fault - can be dispensed with. thus the quality of information in the method proposed here is not impaired even in the event of an error.

According to a development of the invention, it is provided that the at least one metric for each value pair group is normalized to the number of value pairs in the respective value pair group. This improves the robustness of the data analysis. Alternatively or additionally, the plausibility of the at least one metric is checked over different time intervals. Alternatively or additionally, different metrics are preferably checked for plausibility against one another. For example, an alarm is preferably only issued if a sufficient number of value pairs have been measured per value pair group, or the number of value pairs per value pair group does not differ too much from the number of value pairs in the same area in the previous time interval or in other time intervals. By means of the plausibility check, false alarms that arise, for example, due to atypical operation of the internal combustion engine, for example during maintenance or during test runs, can be avoided.

As already stated, within the framework of the method proposed here, an evaluation can take place locally, in particular on a control unit of the internal combustion engine, but also centrally, in particular by means of remote maintenance.

Using the at least one metric, in particular the at least one measurement figure, and / or using the at least one comparison value, errors in other objects or elements of the internal combustion engine are also preferably detected, in particular those errors which in turn contribute indirectly to a reduction or decrease in injector functionality, for example because they contribute to the contamination or wear of the injectors. Here, too, there are mostly characteristic patterns in the value pair groups and in particular metrics.

The object is finally also achieved by creating an internal combustion engine which has a plurality of combustion chambers, with at least one injector being assigned to each combustion chamber. The internal combustion engine also preferably has one exhaust gas temperature sensor per combustion chamber, which is arranged and set up to record exhaust gas temperature values specific to the combustion chamber. The internal combustion engine also has a control unit which is operatively connected to the exhaust gas temperature sensors. The control device is preferably also operatively connected to the injectors. The control unit is set up in particular to carry out a method according to the invention or one of the previously described preferred embodiments of the method for monitoring the injectors of the internal combustion engine.

In connection with the internal combustion engine, there are in particular the advantages that have already been explained in connection with the method.

The description of the method on the one hand and the internal combustion engine on the other hand are to be understood in particular as complementary to one another. Features of the internal combustion engine that have been explicitly or implicitly explained in connection with the method are preferably individually or combined features of a preferred exemplary embodiment of the internal combustion engine. Method steps that have been explained explicitly or implicitly in connection with the internal combustion engine are preferably steps of a preferred embodiment of the method, individually or combined with one another. The method is preferably characterized by at least one step which is caused by at least one feature of an inventive or preferred exemplary embodiment of the internal combustion engine. The internal combustion engine is preferably distinguished by at least one feature which is caused by at least one step of an inventive or preferred embodiment of the method.

• 1 eine schematische Darstellung eines Ausführungsbeispiels einer Brennkraftmaschine;
• 2 eine schematische Darstellung einer Ausführungsform eines Verfahrens zum Überwachen von Injektoren der Brennkraftmaschine, und
• 3 eine Erläuterung der Funktionsweise des Verfahrens anhand verschiedener Diagramme.

The invention is explained in more detail below with reference to the drawing. Show:

• 1 a schematic representation of an embodiment of an internal combustion engine;
• 2 a schematic representation of an embodiment of a method for monitoring injectors of the internal combustion engine, and
• 3 an explanation of how the method works using various diagrams.

1 shows a schematic representation of an embodiment of an internal combustion engine 1 that have a plurality of combustion chambers 3 having each combustion chamber 3 at least one injector 5 , here exactly one injector 5 , assigned. Also, every combustion chamber is 3 an exhaust temperature sensor 7th assigned. The internal combustion engine 1 has a control unit 9 on that with the exhaust gas temperature sensors 7th and preferably with the injectors 5 is operatively connected, in particular on the one hand for reading out the exhaust gas temperature sensors 7th and preferably, on the other hand, to control the injectors 5 .

The control unit 9 is particularly set up around the injectors 5 to monitor. It is set up in particular to carry out the embodiments of a method for monitoring the injectors, which are described in more detail below 5 .

2 shows a schematic representation of such an embodiment of the method. In a first step S1 Combustion chamber-specific exhaust gas temperature values become at least for a group of combustion chambers 3 the internal combustion engine 1 , preferably for all combustion chambers 3 , detected. In a second step S2 becomes an average exhaust gas temperature value in the first step S1 recorded exhaust gas temperature values over the group of combustion chambers 3 , preferably over all combustion chambers 3 , educated. In a third step S3 normalized exhaust gas temperature values are calculated by adding the in the first step S1 recorded exhaust gas temperature values individually with that in the second step S2 calculated exhaust gas temperature mean value can be offset. In particular, a standardized exhaust gas temperature value is obtained for each exhaust gas temperature value.

The first three steps S1 to S3 are repeated several times within a predetermined time interval.

Then in a fourth step S4 Pairs of values formed by for each repetition of the first three steps S1 to S3 the one in the third step S3 Normalized exhaust gas temperature values formed in each case in the third step S3 the same repetition recorded exhaust gas temperature mean is assigned.

In a fifth step S5 are value pair groups of the value pairs for each combustion chamber 3 the internal combustion engine 1 educated.

In a sixth step S6 the value pair groups are evaluated according to at least one metric, and at least one measure per value pair group is obtained.

Preferably the injectors 5 evaluated on the basis of the measures and / or on the basis of the occurrence of certain value pair groups.

Preferably the first six steps S1 to S6 repeated for a plurality of the predetermined time intervals.

Preferably in a seventh step S7 those in the fifth step S5 Value pair groups formed for different time intervals are assigned to one another individually in the combustion chamber. In an eighth step S8 will be the ones in the sixth step S6 Values of pairs of values assigned to one another, obtained in the different time intervals, are compared with one another, and at least one comparison value is made per group of values, and preferably per combustion chamber 3 - receive.

Preferably the injectors 5 evaluated on the basis of the comparative value.

Functionality of the at least one injector is preferred 5 for every combustion chamber 3 the internal combustion engine 1 evaluated in that the at least one measure is compared with a measure threshold value, and / or in which the at least one comparison value is compared with a comparison value threshold value.

A development over time of the at least one dimension figure and / or of the at least one comparison value is preferably recorded, with the at least one injector 5 for every combustion chamber 3 is evaluated based on the recorded development over time.

A prognosis of the functionality of the at least one injector is preferably made 5 for every combustion chamber 3 determined on the basis of an extrapolation of the recorded development over time.

A value pair density in a plane spanned by the normalized exhaust gas temperature and the average exhaust gas temperature value is preferably used as the grouping criterion for the formation of the value pair groups.

The at least one metric is preferably selected from an in particular weighted center of gravity, a - in particular geometric - center position, an expansion of the value pair group, in particular in the direction of the standardized exhaust gas temperature, a position of a maximum - in particular the standardized exhaust gas temperature -, a value pair density, and a number of Value pairs in the value pair group.

Preferably, only exhaust gas temperature values or normalized exhaust gas temperature values are used which, when recorded, indicate a steady-state operation of the internal combustion engine 1 is present within a predetermined steady-state time interval.

The at least one metric for each value pair group is preferably normalized to the number of value pairs in the respective value pair group. Alternatively or additionally, the at least one metric is checked for plausibility over different time intervals and / or different metrics are checked for plausibility against one another.

3 shows an explanation of the mode of operation of the method proposed here on the basis of various diagrams.

At a) there are pairs of values for a functional injector 5 shown in a diagram that is spanned by the standardized exhaust gas temperature T _norm on the ordinate and the average exhaust gas temperature value T _m on the abscissa. For every combustion chamber 3 the internal combustion engine 1 such a diagram is obtained within a predetermined time interval. Doing the first three steps for each repetition S1 to S3 each receive an average exhaust gas temperature.

When repeating the first six steps S1 to S6 for a plurality of the predetermined time intervals is then in each case for each time interval - and for each combustion chamber 3 - Obtain such a diagram so that an injector has developed over time 5 can be assessed in particular by comparing the various diagrams.

One of the combustion chambers is consequently shown in the diagram at a) 3 associated value pairs, which were measured in one of the predetermined time intervals. The accumulations of the pairs of values in different sections along the abscissa describe different load points of the internal combustion engine 1 , for example on the far left points when idling, on the far right, for example, pairs of values in the vicinity of full load operation. These load points can easily be identified on the basis of the mean exhaust gas temperature values _{T m;} a higher exhaust gas temperature mean value is assigned to full load than to idling operation. Furthermore, the value pairs can have different operating states of the combustion chamber under consideration 3 assigned injector 5 be assigned. For example, describes a first accumulation H1 that are with an oval border is marked, an operation of the injector 5 with the pilot needle fully open, which can be clearly seen from the relatively small spread of the value pairs in the vertical direction.

A second cluster H2 is also marked with an oval border. This second cluster H2 of value pairs describes an operation of the injector 5 in the ballistic area in which the pilot needle does not reach the full stop. This area corresponds to a characteristic kink in the injector curve.

• Es werden Bereiche auf der horizontalen und gegebenenfalls vertikalen Achse definiert, beispielsweise ein erster Bereich zwischen 180 °C und 350 °C und ein zweiter Bereich zwischen 450 °C und 550 °C - jeweils auf der Abszisse -, wobei die Bereiche wahlweise auch beschrieben werden können durch beliebige nicht-lineare Grenzkurven in der Diagrammebene. Alternativ können Punktewolken durch einen Klassifikationsalgorithmus bestimmt werden, die insbesondere eine gewisse Mindest-Punktedichte - das heißt Anzahl von Wertepaaren pro Fläche in der Diagrammebene - aufweisen. Diese Punktewolken oder Wertepaargruppen werden bevorzugt durch ihre Lage beschrieben, beispielsweise durch die Lage des Mittelpunkts und/oder eine Ausdehnung bezüglich der beiden Diagrammachsen. Im Fall von Diagramm a) würde ein solcher Algorithmus beispielsweise drei bis fünf solcher Wertepaargruppen identifizieren: Eine kleine bei 100 °C, bis zu drei Wolken zwischen 180 °C und 320 °C, und eine weitere Wertepaargruppe bei mehr als 450 °C auf der Abszisse.

Metrics can now be created for certain areas of the diagram, for example as follows:

• Ranges are defined on the horizontal and possibly vertical axis, for example a first range between 180 ° C and 350 ° C and a second range between 450 ° C and 550 ° C - in each case on the abscissa - with the ranges optionally also being described can through any non-linear limit curves in the diagram level. Alternatively, point clouds can be determined by a classification algorithm, which in particular have a certain minimum point density - that is, number of value pairs per area in the diagram plane. These point clouds or value pair groups are preferably described by their position, for example by the position of the center point and / or an extension with respect to the two diagram axes. In the case of diagram a), such an algorithm would identify three to five such value pair groups, for example: A small one at 100 ° C, up to three clouds between 180 ° C and 320 ° C, and a further value pair group at more than 450 ° C on the Abscissa.

Metrics can now be defined for each of the areas mentioned, i.e. for each value pair group. A metric is, for example, an extent in the direction of the ordinate. Another metric can be the center of gravity of the points; another the density of the dots; also the number of points per area. Of course, other metrics are conceivable.

As already stated, by repeating the first six steps S1 to S6 different such diagrams are formed for a plurality of the predetermined time intervals. In particular, arises for every time interval - and for every combustion chamber or injector 5 - a characteristic diagram.

The metrics of the same areas, that is, value pair groups from different time intervals, are now compared with one another. In the case of an undamaged injector, it is expected that, for example, the expansion in the vertical direction does not change appreciably and that the overall pattern remains very similar as long as the internal combustion engine is used 1 is operated in a similar manner as in the previous time interval, which is usually the case - especially with a suitable choice of the time interval. This has also been shown in numerous evaluations of engine data from the field.

Even small previous damage to injectors 5 however - as numerous measurements in the field have also shown - lead to significant changes in the pattern and usually to a significant increase in the extent of the value pair groups in the direction of the ordinate, or to a shift in the center of gravity. It could also be shown that the previous damage to an injector 5 can be proven in the diagrams weeks and sometimes months before a classic alarm is triggered.

Depending on the type of damage to an injector 5 there are very specific changes in the metrics. For various injector damage, metrics change in very specific areas, for example preferably in the area which can be assigned to the open pilot needle in the event of damage to the nozzle of the injector 5 , for example by washing out the nozzle cross-section; or preferably in the area that can be assigned to the ballistic area of the pilot needle for damage in the hydraulic system of the injector. Furthermore, certain metrics change in a characteristic way for various injector damage, for example there is a shift and an increase in the value pair groups towards higher values along the ordinate for damage to the seat of the pilot needle; or there is an occurrence of value pairs with a very low point density in a region of high values on the ordinate with a very large extent of the corresponding point cloud for a pilot needle jammed in its guide.

In order to detect injector damage, the metrics and their changes to the previous time interval can now be monitored. If the metrics exceed defined threshold values, specific alarms can be issued that indicate certain injector damage. In addition, the degree of previous damage to an injector can be derived from the metrics 5 can be estimated, and there can possibly be a failure probability for the injector concerned 5 be derived. If the metrics are observed over different time intervals, the course of the damage can be tracked and, if necessary, extrapolated, i.e. predicted. In this way, it is possible to identify suitable measures at an early stage and to better plan outages and maintenance.

It is also possible to check the plausibility of the metrics with respect to one another, for example by only issuing an alarm if enough points have been measured per value pair group or if the number of points per value pair group does not differ too much from the number of points in the same area in the previous time interval or in other time intervals. In this way, false alarms that could arise, for example, due to atypical engine operation - for example during maintenance or test runs - can be avoided. Finally, wherever necessary, the metrics can be viewed standardized to the number of points per value pair group, which improves the robustness of the analysis of the data.

At b) is the diagram for the injector 5 of the diagram from a) for a later time interval. This can result in pre-existing damage to the injector 5 be recognized. With H2 the original position of the second cluster is designated here once again, as it resulted in the diagram at a). The new position is marked with H2 '.

It can now be seen at b) that the extension and position of the points in the range between 180 ° C and 350 ° C on the abscissa has clearly changed towards higher values. In addition, a clear change in the center of gravity of the value pair group between 450 ° C and 550 ° C can be observed, where again the original position of the first cluster for comparison H1 is marked according to the diagram at a) in the diagram for b). For the plausibility check, it is also relevant that - which is not shown here - similar changes in the other injectors 5 not to be observed, which with increased certainty of a previous damage to the injector considered here 5 closes.

A third cluster H3 is again outlined in an oval. This occurs here in a range from 200 ° C to 250 ° C on the abscissa. This third cluster H3 indicates a changed behavior of the injector in its ballistic area. From the evaluation of a large number of data for a large number of internal combustion engines 1 and injectors 5 it is known that such a change indicates damage in the area of the pilot seat, in particular scratches in the seat, in particular a resulting leak.

At c) is a diagram showing an example of another type of damage: Here falls, a few weeks before the injector fails 5 , in addition to a significant previous damage similar to the example according to the diagram at b) a fourth accumulation H4 of pairs of values with a very low point density and high values with respect to the ordinate between 300 ° C and 400 ° C on the abscissa, with this fourth cluster H4 marked with an oval border. These values indicate a sporadic fault in the hydraulic system of the injector 5 close with a strong expression, which is largely due to the regular behavior of the injector, which has already been damaged 5 alternates. This type of error pattern can be derived from the evaluation of a large number of data for a large number of internal combustion engines 1 on a sporadic jamming of the pilot needle in its guide, which is caused by particles.

The diagrams can preferably also be used to identify errors in other objects or elements of the internal combustion engine 1 getting closed.

Claims (11)

A method for monitoring injectors (5) of an internal combustion engine (1) with a plurality of combustion chambers (3), at least one injector (5) being assigned to each combustion chamber (3), with the following steps: a) Acquisition of combustion chamber-specific exhaust gas temperature values for at least one Group of combustion chambers (3) of the internal combustion engine (1); b) Forming an average exhaust gas temperature value (T _m ) of the exhaust gas temperature values recorded in step a) over the group of combustion chambers (3); c) calculating normalized exhaust gas temperature values (T _norm ) by individually calculating the exhaust gas temperature values recorded in step a) with the average exhaust gas temperature value (T _m ) formed in step b); d) repeating steps a) to c) multiple times within a predetermined time interval; _{e) Forming pairs of values by assigning the average exhaust gas temperature value (T m} ) recorded in step b) of the same repetition to the _{normalized exhaust gas temperature values (T norm} ) formed in step c) for each repetition of steps a) to c); f) Forming value pair groups of the value pairs for each combustion chamber (3) of the internal combustion engine (1), and g) evaluating the value pair groups according to at least one metric and obtaining at least one measure per value pair group. Procedure according to Claim 1 , characterized in that h) steps a) to g) are repeated for a plurality of the predetermined time intervals. Procedure according to Claim 2 , characterized in that i) the value pair groups formed for different time intervals in step f) each other can be assigned individually to the combustion chamber, wherein preferably j) the measured values of value pair groups assigned to one another are compared in the various time intervals in each case in step g), with at least one comparison value per value pair group being obtained. Procedure according to Claim 3 , characterized in that the functionality of the at least one injector (5) is evaluated for each combustion chamber (3) of the internal combustion engine (1) by comparing the at least one measure with a measure threshold value, and / or the at least one comparison value is compared with a comparison value threshold value. Method according to one of the Claims 3 or 4th , characterized in that a development over time - of the at least one measure and / or - of the at least one comparison value is recorded, the at least one injector (5) for each combustion chamber (3) being evaluated on the basis of the recorded development over time. Procedure according to Claim 5 , characterized in that a prognosis of the functionality of the at least one injector (5) is recorded for each combustion chamber (3) on the basis of an extrapolation of the recorded development over time. Method according to one of the preceding claims, characterized _{in that a value pair density in a plane spanned} by the normalized exhaust gas temperature (T norm) and the average exhaust gas temperature _{value (T m} ) is used as the grouping criterion for the formation of the value pair groups in step f). Method according to one of the preceding claims, characterized in that the at least one metric is selected from a group consisting of: a center of gravity, a center position, an extension, a position of a maximum, a value pair density, and a number of value pairs in the value pair group. Method according to one of the preceding claims, characterized in that only exhaust gas temperature _{values or standardized exhaust gas temperature values (T norm} ) are used which, when recorded, indicate that the internal combustion engine (1) is in steady-state operation within a predetermined steady-state time interval. Method according to one of the preceding claims, characterized in that - the at least one metric for each value pair group is normalized to the number of value pairs in the respective value pair group; and / or that - the plausibility of the at least one metric is checked over different time intervals; and / or that different metrics are checked for plausibility against one another. Internal combustion engine (1) with a plurality of combustion chambers (3), each combustion chamber (3) of the plurality of combustion chambers (3) being assigned at least one injector (5), each combustion chamber (3) of the plurality of combustion chambers (3) being assigned an exhaust gas temperature sensor (7) is assigned, and with a control unit (9) which is operatively connected to the exhaust gas temperature sensors (7), the control unit (9) being set up to carry out a method according to one of the Claims 1 to 10 .

Images